Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 1 of 112
`
`
`
`
`
`
`
`
`
`
`
`
`
`IN THE UNITED STATES DISTRICT COURT
`
`FOR THE NORTHERN DISTRICT OF CALIFORNIA
`
`CITY AND COUNTY OF SAN
`FRANCISCO, et al.,
`
`Plaintiffs,
`
`v.
`
`PURDUE PHARMA L.P., et al.,
`
`Defendants.
`
`
`
`
`
`
`
`Case No. 18-cv-07591-CRB
`
`
`FINDINGS OF FACT AND
`CONCLUSIONS OF LAW
`REGARDING WALGREENS
`
`The opioid epidemic has plagued San Francisco for over twenty years. The number
`
`of individuals who die annually from opioid overdoses continues to climb. Thousands of
`
`city residents, from all walks of life, struggle with addiction. Widespread opioid use has
`
`strained the city’s hospitals. It has forced streets, parks, and public spaces to close. It has
`
`exacerbated crime and homelessness. Every year, San Francisco devotes significant
`
`resources to a multiprong fight against the opioid epidemic. That fight includes this case.
`
`This case is part of a nationwide multidistrict litigation stemming from the ongoing
`
`opioid epidemic. Cities, counties, and states across the country have filed claims against
`
`manufacturers, distributors, and dispensers of prescription opioids. While the facts of each
`
`case vary, the claims center on the contention that each defendant has contributed to the
`
`opioid epidemic that has engulfed the country.
`
`
`
`In this case, the People of the State of California, acting through the San Francisco
`
`City Attorney (“Plaintiff”), filed claims against dozens of defendants related to the opioid
`
`epidemic in San Francisco. By the time of trial, only four defendants remained. The Court
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 2 of 112
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`held a bench trial from April 25, 2022 to June 27, 2022. Closing argument was held from
`
`July 12 to July 13, 2022. By the close of trial, Walgreens Co. (“Walgreens”) was the sole
`
`remaining defendant. The other three defendants settled their claims.
`
`
`
`At trial, Plaintiff brought a single public nuisance claim against Walgreens. The
`
`question for the Court is whether Plaintiff proffered sufficient evidence at trial to prove
`
`this claim. To carry its burden of proof, Plaintiff had to establish that it is more likely than
`
`not that Walgreens knowingly engaged in unreasonable conduct that was a substantial
`
`factor in contributing to the opioid epidemic in San Francisco. After careful consideration
`
`of the evidence, the Court finds that Plaintiff carried its burden.
`
`
`
`Walgreens is the largest retail pharmacy chain in San Francisco. Between 2006 and
`
`2020, Walgreens distributed and dispensed over one hundred million prescription opioid
`
`pills in the city. The Controlled Substances Act (“CSA”) and its implementing regulations
`
`impose duties on distributors and dispensers of prescription opioids. In exchange for the
`
`privilege of distributing and dispensing prescription opioids, Walgreens has regulatory
`
`obligations to take reasonable steps to prevent the drugs from being diverted and harming
`
`the public. The evidence at trial established that Walgreens breached these obligations.
`
`
`
`Until 2014, Walgreens distributed prescription opioids to its pharmacies in San
`
`Francisco. CSA regulations require distributors to implement and maintain a system for
`
`identifying suspicious orders of opioids. Suspicious orders of opioids must be halted and
`
`reported to the DEA. They cannot be shipped to the ordering pharmacy. The evidence at
`
`trial established that Walgreens violated this regulatory duty for several years. It did not
`
`maintain an effective system for identifying suspicious orders. It shipped thousands of
`
`suspicious orders to its pharmacies without investigation. In 2012, the DEA shut down
`
`one of Walgreens’ three controlled substance distribution centers because the distribution
`
`center’s failure to monitor for suspicious opioid orders posed an imminent threat of harm
`
`to public health and safety. Shortly thereafter, Walgreens stopped distributing opioids all
`
`together.
`
`Walgreens pharmacies are the largest dispenser of opioids in San Francisco. To
`
`2
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 3 of 112
`
`
`
`prevent diversion, CSA regulations require Walgreens to verify the medical legitimacy of
`
`opioid prescriptions before dispensing them. Fulfilling this duty requires Walgreens
`
`pharmacies to resolve “red flags” associated with a prescription before dispensing it. Red
`
`flags are well-established warning signs that raise questions about the legitimacy of a
`
`prescription. Medically legitimate prescriptions are prescribed for a patient’s benefit, but
`
`medically illegitimate prescriptions are not. They are prescriptions that are misused and
`
`abused. Medically illegitimate prescriptions extend far beyond forged prescriptions and
`
`prescriptions that are written on a stolen prescription pad. Many illegitimate prescriptions
`
`come from unscrupulous doctors who write prescriptions in exchange for payment. It is
`
`not enough for a pharmacy to simply ascertain that a licensed prescriber wrote the
`
`prescription. Pharmacies have a corresponding duty to exercise independent judgment in
`
`determining whether the prescription was written for a legitimate medical purpose.
`
`The evidence at trial established that from 2006 to 2020, Walgreens pharmacies in
`
`San Francisco dispensed hundreds of thousands of red flag opioid prescriptions without
`
`performing adequate due diligence. Tens of thousands of these prescriptions were written
`
`by doctors with suspect prescribing patterns. The evidence showed that Walgreens did not
`
`provide its pharmacists with sufficient time, staffing, or resources to perform due diligence
`
`on these prescriptions. Pharmacists experienced constant pressure to fill prescriptions as
`
`quickly as possible, and a shortage of resources to review them before dispensing. As a
`
`result of Walgreens’ fifteen-year failure to perform adequate due diligence, Plaintiff
`
`proved that it is more likely than not that Walgreens pharmacies dispensed large volumes
`
`of medically illegitimate opioid prescriptions that were diverted for illicit use and that
`
`substantially contributed to the opioid epidemic in San Francisco.
`
`
`
`The Court’s findings of fact and conclusions of law are set forth below. This ruling
`
`holds only that Walgreens is liable for substantially contributing to the public nuisance in
`
`San Francisco. A subsequent trial will determine the extent to which Walgreens must
`
`abate the public nuisance that it helped to create.
`
`3
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 4 of 112
`
`
`
`I.
`
`FINDINGS OF FACT
`
`In a bench trial, the court’s findings of fact are presumed to be based on admissible
`
`evidence. Williams v. Illinois, 567 U.S. 50, 69 (2012); Harris v. Rivera, 454 U.S. 339, 465
`
`(1981). To the extent that objections have been raised to the evidence cited in support of
`
`the Court’s findings, the objections are overruled. See City of Huntington v. Amerisource
`
`Bergen Drug Corp., No. CV 3:17-01362, 2022 WL 2399876, at *1 (S.D.W. Va. July 4,
`
`2022).
`
`A.
`
`Background
`
`
`
`The Science of Opioid Addiction
`
`Opioid addiction is explained by a change in an opioid user’s brain chemistry.1 See
`
`Lembke (dkt. 1281) Decl. ¶ 3. Opioids bind to mu-pain receptors temporarily relieving
`
`pain. Lembke, May 9, 2022, Trial Tr. at 383:23–384:4. In addition, opioids cause the
`
`release of dopamine. Id. Dopamine is a naturally occurring neurotransmitter that causes
`
`feelings of pleasure and reward. See id.; Lembke, May 9, 2022, Trial Tr. at 383:23–384:9.
`
`In response to repeated additional releases of dopamine from opioid use, the brain begins
`
`to downregulate the amount of dopamine it naturally produces, a process known as
`
`neuroadaptation. Lembke Decl. ¶ 3. The result is a dopamine deficient state, in which the
`
`brain is producing less dopamine and the user is experiencing less pleasure and more pain
`
`than they were before opioid use began. Id.
`
`In a dopamine deficient state, a user needs opioids to return to their previous
`
`dopamine baseline and to avoid the pain of prolonged dopamine deficiency. Lembke Decl.
`
`¶ 3. Users “need opioids not to feel good but just to restore a level balance and feel
`
`normal.” See Lembke, May 9, 2022, Trial Tr. at 384:10–386:3. Opioid users in this state
`
`are physically dependent on the drugs. See id. at 384:10–386:15. Someone taking opioids
`
`
`1 Opioid addiction is synonymous with opioid use disorder. See Colwell, April 28, 2022, Trial Tr.
`at 360. Opioid use disorder has a more precise medical definition set out in The Diagnostic and
`Statistical Manual of Mental Disorders (“DSM”), which defines the severity as mild, moderate, or
`severe, depending on the symptoms present. See Lembke Decl. ¶ 2. But both terms describe the
`same form of harmful behavior: the continued use of opioids despite deleterious effects to self or
`others. See id. ¶¶ 2–5.
`
`4
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 5 of 112
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`for “relatively short periods of time” can develop physical dependence and experience
`
`withdrawal if they stop taking opioids. Zevin, May 10, 2022, Trial Tr. at 640:21–641:1;
`
`Coffin (dkt. 1376) Decl. ¶ 55. Symptoms of withdrawal include anxiety, debility,
`
`insomnia, dysphoria, “and in the case of opioids, a very distinct and painful physical
`
`withdrawal syndrome, including full-body pain that can be experienced and is typically
`
`experienced in people who do not have a pain disorder.” Lembke, May 9, 2022, Trial Tr.
`
`at 384:10–386:15.
`
`Over time, opioid users generally require higher doses to experience the same effect
`
`that they initially experienced. Lembke, May 9, 2022, Trial Tr. at 386:12–387:15. This is
`
`the process of developing tolerance to the drug. Id. Dr. Lembke explained that “the brain
`
`adapts to the presence of the opioid molecule such that the individual needs more and more
`
`to get the same effect and ultimately is physically dependent and experiences painful
`
`withdrawal when they stop whether or not they have a pain condition.” Id. at 387:11–15.
`
`Once the brain adapts to the presence of opioids, it can “take a very long time after the
`
`individual has stopped using their drug for the brain to reset itself to normal dopamine
`
`levels.” Id. at 387:17–388:7; 391:25–392:18. Reducing opioid use requires tapering,
`
`which involves gradually progressing to lower doses of opioids. Lembke Decl. ¶¶ 39–41.
`
`The process of tapering off opioids “is time-intensive and requires substantial support from
`
`clinicians and other providers in the healthcare system.” Coffin Decl. ¶ 55.
`
`Opioid addiction affects people from all walks of life, regardless of age, ethnicity,
`
`or socioeconomic status. Colwell (dkt. 1284) Decl. ¶¶ 11–12. The neural pathways
`
`affected by opioid use are common across all people, which makes everyone vulnerable to
`
`opioid addiction. Lembke Decl. ¶ 4. Opioids carry risks of addiction even when
`
`prescribed by a medical professional. Keyes (dkt. 1386) Decl. ¶ 14, 40. Stronger dosages
`
`and longer durations of use increase the risk of addiction. Lembke, May 9, 2022, Trial Tr.
`
`at 398:5–16.
`
`Opioid addiction can have devastating consequences. People suffering from the
`
`most severe forms of addiction “commit all available resources to obtaining more of the
`
`5
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 6 of 112
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`substance, even forgoing natural rewards like food, finding a mate, or raising children.”
`
`Lembke Decl. ¶ 5. For people suffering from severe addiction, consuming more opioids
`
`often appears to be the only way to avoid the intense pain of withdrawal. See id.
`
`
`
`The Opioid Epidemic in San Francisco
`
`For over two decades, San Francisco has been battling an opioid epidemic, defined
`
`by high rates of opioid abuse and addiction throughout the city.2 Coffin Decl. ¶ 11. The
`
`number of people in the city abusing opioids has substantially accelerated in recent years.
`
`Zevin Decl. (dkt. 1296) ¶ 5. The number of opioid-related emergency room visits has
`
`increased significantly, more than tripling from 886 in 2015 to 2,998 in 2020. Coffin Decl.
`
`¶ 40. Since 2016, opioid overdoses have been the leading cause of death among the
`
`homeless in San Francisco. See Zevin Decl. ¶ 6. In 2019, the last year of available data,
`
`an estimated 40,958 city residents out of a total population of approximately 865,000
`
`suffered opioid addiction. See Keyes Decl. ¶ 25. That same year, approximately 1,939
`
`people in San Francisco overdosed on opioids, an average of 5.3 opioid overdoses per day.
`
`Id. ¶ 135.
`
`Prescription opioids have been at the heart of San Francisco’s ongoing opioid
`
`epidemic, which has unfolded in three different waves. The first wave started in the late
`
`1990s and early 2000s when opioid manufacturers began to aggressively promote opioids
`
`as safe and effective for treating a broad range of medical conditions. Lembke, May 9,
`
`2022, Trial Tr. at 382:15–383:11. The aggressive marketing resulted in increased
`
`prescribing and the increased prescribing resulted in an increase in opioid abuse, addiction,
`
`and overdoses. See id.; Coffin Decl. ¶¶ 17–19; 37. The second wave began in the early
`
`2010s, when medical professionals began to reduce opioid prescribing based on the
`
`recognition that opioids are not a safe and effective form of treatment for many medical
`
`conditions. See id. ¶¶ 20–22. However, the massive expansion of the prescription opioid
`
`
`2 “Epidemic” describes the outbreak of a disease that spreads quickly, affecting many people at the
`same time. See Lembke Decl. Part III.Q. The term appropriately describes the rapid increase in
`opioid-related morbidity and mortality that began in the 1990s and continues still. See id.
`6
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 7 of 112
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`supply that happened in the preceding decade laid the foundation for increased use of illicit
`
`opioids. Keyes Decl. ¶ 12. In the second wave, many people who were addicted to
`
`prescription opioids but no longer readily able to obtain them from doctors shifted to
`
`heroin use. See id.; Coffin Decl. ¶ 37. The third wave started around 2015, when
`
`inexpensive and highly potent fentanyl became widely available across a city already
`
`struggling with opioid addiction. Keyes Decl. ¶ 12; Coffin Decl. ¶ 37. The city’s fight
`
`against the effects of each wave of the opioid epidemic continues today.
`
`a. Wave One: Prescription Opioids
`
`Heroin use has long been a public health issue in San Francisco. See Coffin Decl. ¶
`
`12. In the late 1990s, San Francisco pioneered public health programs to address heroin
`
`use in the city, including expanding the availability of treatment programs and resources.
`
`Id. ¶¶ 12–14; Coffin, May 26, 2022, Trial Tr. at 1896:18–1897:16. The public health
`
`programs worked. By the early 2000s, the city was winning the battle against heroin use.
`
`See Coffin, May 26, 2022, Trial Tr. at 1896:18–1897:16. Heroin overdoses decreased
`
`dramatically from 150 per year in the late 1990s to 10 in 2010. Id.
`
`As San Francisco was succeeding in the fight against heroin use, the volume of
`
`prescription opioids began to increase significantly. See Coffin Decl. ¶¶ 11, 16–17.
`
`The significant increase in opioid prescribing that took place in the late 1990s throughout
`
`the 2000s resulted from changing views on the safety and efficacy of opioids as a form of
`
`pain treatment. Lembke, May 9, 2022, Trial Tr. at 382:15–19; Zevin Decl. ¶ 9. For much
`
`of the twentieth century, medical professionals used opioids sparingly because of “the
`
`legitimate concern that patients would get addicted.” Lembke Decl. ¶ 7; see also Herzberg
`
`Decl. ¶¶ 4–5. Opioid use was generally limited to treating cancer pain, hospice patients,
`
`and for short-term use in treating pain stemming from severe injuries. See Zevin, May 10,
`
`2022, Trial Tr. at 620:17–621:25. Medical professionals understood that opioids carried
`
`high risks of abuse, and they limited their prescribing. See Lembke, May 9, 2022, Trial Tr.
`
`at 394:7–13. This “conservative consensus” about opioid prescribing held for decades,
`
`until it began to change in the 1990s. Herzberg, May 31, 2022, Trial Tr. at 1975:18–
`
`7
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 8 of 112
`
`
`
`1976:12
`
`In the 1990s and throughout the 2000s, opioid manufacturers launched aggressive
`
`marketing campaigns intended to persuade medical professionals that prescription opioids
`
`were a safe and effective form of treatment for many pain conditions. Lembke Decl. ¶¶
`
`11–12; Herzberg Decl. ¶ 11; Coffin Decl. ¶ 17. The marketing campaigns aimed to
`
`expand the market for prescription opioids from a limited range of acute conditions to
`
`commonplace forms of pain, such as lower back pain and headaches. Lembke Decl. ¶¶
`
`11–12; Herzberg Decl. ¶ 11. The marketing campaigns claimed that new formulations of
`
`prescription opioids provided highly effective pain relief and could be safely prescribed for
`
`long-term use with low risks of addiction. See Lembke, May 9, 2022, Trial Tr. at 393:14–
`
`395:13; Herzberg, May 31, 2022, Trial Tr. at 1982:9–1987:12; Keyes Decl. ¶¶ 91–99. The
`
`marketing campaigns were extensive. Opioid manufacturers employed large salesforces,
`
`targeted high-prescribing doctors, held medical conferences promoting opioids for pain
`
`treatment, paid doctors to speak about the benefits of opioids, and funded articles in
`
`medical journals that highlighted the safety and efficacy of opioids. Lembke Decl. ¶¶ 50–
`
`97; Herzberg Decl. ¶ 13. The marketing campaigns worked. The aggressive promotion of
`
`the use of prescription opioids to treat widespread forms of pain led to a paradigm shift in
`
`the treatment of pain. Opioids became “first-line treatment for minor and chronic pain
`
`conditions.” Lembke Decl. ¶ 13; Lembke, May 9, 2022, Trial Tr. at 382:5–5.
`
`In addition to marketing opioids as a safe and effective form of treatment, opioid
`
`manufacturers broadly promoted the idea that pain was commonplace and undertreated.
`
`See Lembke Decl. ¶ 10. Opioid manufacturers represented that millions of people across
`
`the country were living with chronic pain that affected their qualify of life. See Lembke,
`
`May 9, 2022, Trial Tr. at 393:14–395:13. Pain became the “fifth vital sign.” Lembke
`
`Decl. ¶ 103. Patients began using a visual pain scale that consisted of “a series of happy or
`
`sad faces supposedly corresponding to pain levels from 0 (no pain) to 10 (the most extreme
`
`pain)” to communicate their pain to doctors. Id. State Medical Boards that received
`
`significant support and funding from opioid manufacturers warned doctors that failing to
`
`8
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 9 of 112
`
`
`
`adequately treat pain could be grounds for a malpractice claim. See id. ¶¶ 90–96. Doctors
`
`“lived in fear of disciplinary action from the State Medical Boards and the lawsuit that
`
`usually followed, if they denied a patient opioid painkillers.” Id. ¶ 93.
`
`Opioid manufacturers billed prescription opioids as the solution to undertreated
`
`pain. See Lembke Decl. ¶ 13; Lembke, May 9, 2022, Trial Tr. at 384:21–395:7. Doctors
`
`across all specialties began prescribing more opioids to treat a broad range of conditions.
`
`Lembke Decl. ¶ 10; Keyes Decl. ¶¶ 88–89. Drug manufacturers claimed that there was no
`
`“ceiling dose” for opioids, and doctors began writing prescriptions for higher dosages and
`
`longer durations. Lembke Decl. Part C. The number of prescriptions increased markedly.
`
`Coffin Decl. ¶¶ 17–18; Coffin, May 26, 2022, Trial Tr. at 1912:19–1913:6. From the late
`
`1990s to 2012, opioid prescribing quadrupled. Lembke Decl. ¶ 8.
`
`The massive increase in opioid prescriptions caused corresponding increases in
`
`opioid abuse, addiction, and overdoses. Keyes Decl. ¶¶ 11–13; Lembke ¶¶ 8–9 (“Higher
`
`doses and longer durations cause increased rates of addiction and death.”). The most
`
`important risk factor for opioid addiction is access to opioids.3 Lembke Decl. ¶ 4. As the
`
`opioid supply in San Francisco increased, the rates of opioid addiction and overdoses
`
`increased as a “direct consequence.” Keyes Decl. ¶¶ 11–13. Each “one-pill increase in per
`
`capita pill volume” is “associated with 0.2 additional overdose deaths.” Id. ¶ 79. In
`
`addition to the harms resulting directly from the increasing supply, many of the claims that
`
`manufacturers made about opioids proved to be inaccurate. See Lembke Decl. Part III. As
`
`rates of abuse and addiction increased, it became clear that opioids were not a safe and
`
`effective form of treatment for many forms of pain. See id.
`
`By the early 2000s, prescription opioids had surpassed heroin as the leading cause
`
`of opioid overdoses in San Francisco. See Coffin, May 26, 2022, Trial Tr. at 1898:4–16.
`
`From 2000 to 2010, prescription opioids caused the overwhelming majority of opioid
`
`
`3 A risk factor is a variable that increases the frequency with which an outcome occurs. Keyes
`Decl. ¶ 31. An example is cigarette smoking and lung cancer. Cigarette smoking increases the
`risk of lung cancer (especially the longer and more frequently that someone smokes). Id.
`9
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 10 of 112
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`overdose deaths in San Francisco. See Coffin Decl. ¶ 19; Coffin, May 26, 2022, Trial Tr.
`
`at 1906:6–20. Overdose deaths from prescription opioids continued to increase despite the
`
`city’s significant investment in public health programs designed to combat opioid abuse.
`
`Coffin Decl. ¶ 28. From 2010 to 2012, the rate of opioid overdoses in San Francisco was
`
`2.23 times the national average. Id. ¶ 21.
`
`In the early 2010s, the medical community’s view on prescription opioids began to
`
`change again. See Coffin Decl. ¶¶ 20–22. Rates of opioid abuse, addiction, and overdoses
`
`had increased across the country for a decade. See P-19656 at 00009 (2012 DEA
`
`presentation to pharmaceutical companies stating “more Americans abuse prescription
`
`drugs than the number of: cocaine, hallucinogens, heroin, and inhalant abusers
`
`combined!!!”) (cleaned up). There was a growing spotlight on recreational use of
`
`prescription opioids, including abuse by teenagers and adolescents. See id. at 00011,
`
`00014, 00020. In addition, there was growing awareness that the medical profession and
`
`government regulators had underappreciated opioids’ risks of addiction. See Zevin Decl.
`
`¶¶ 8–10; Coffin, May 26, 2022, Trial Tr. at 1899:7–14. Medical professionals began to
`
`recognize that the “overreliance on opioids” to treat pain had caused “a crisis nationally
`
`and in San Francisco.” Coffin, May 26, 2022, Trial Tr. at 1899:7–14. Doctors began to
`
`take a more careful approach to opioid prescribing and opioid prescribing rates began to
`
`decline. See Coffin, May 26, 2022, Trial Tr. at 1899:7–20; Coffin Decl. ¶¶ 20–22.
`
`b. Wave Two: Prescription Opioids and Heroin
`
`The volume of opioids prescribed in San Francisco peaked in the early 2010s. See
`
`Coffin Decl. ¶ 11, 22. Opioid prescribing rates declined from 2011 to 2020, although they
`
`remained significantly higher than pre-2000s levels. Keyes Decl. ¶ 22. As prescription
`
`opioids became less readily available, many people addicted to prescription opioids began
`
`to turn to heroin. While the city had a heroin problem in the late 1990s, the problem
`
`became “significantly worse” following the increase in opioid prescribing in the 2000s.
`
`See Tong Decl. (dkt. 1336) ¶ 3; Tong, May 19, 2022, Trial Tr. at 1351:22–1352:6.
`
`In the early 2010s, heroin use and heroin overdose deaths increased as the volume
`
`10
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 11 of 112
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`of prescription opioids decreased. See Coffin, May 26, 2022, Trial Tr. at 1899:15–1900:2;
`
`Coffin Decl. ¶ 24 (“[A]s prescription opioid use and prescribing decreased in SF, heroin
`
`use began to increase again.”). The massive increase in prescription opioids in the 2000s
`
`helped to increase the demand for heroin in the 2010s. See Coffin Decl. ¶ 24; Keyes Decl.
`
`¶ 12. The link between prescription opioid use and heroin use is well established. See
`
`Coffin Decl. ¶¶ 24–27; see also P-19656 at 00051. The “most significant risk factor for
`
`heroin use is exposure to prescription opioids.” Keyes Decl. ¶¶ 53, 66–70. Multiple
`
`studies have found that approximately 70–80% of heroin users in the last two decades used
`
`prescription opioids before using heroin. Id. ¶ 18. Approximately 75% of people who
`
`suffer an opioid overdose began their opioid use with prescription opioids. Id. ¶ 10.
`
`People who stop using prescription opioids have an increased risk of transitioning to
`
`heroin. Id. ¶ 107. From 2002 to 2013, heroin use increased 138% among those who use
`
`prescription opioids. Id. ¶ 57. Like opioid abuse, heroin use affects people from all walks
`
`of life. In recent years, heroin use has increased across race, gender, and social class. See
`
`id. ¶ 56.
`
`
`
`Replacement theory and the gateway effect explain the transition from prescription
`
`opioids to heroin use. See Lembke Decl. ¶¶ 153–156; Coffin, May 26, 2022, Trial Tr. at
`
`1899:11–1900:7. Replacement occurs when people are unable to obtain opioids from a
`
`medical professional and resort to heroin use to avoid the pains of withdrawal. See id. As
`
`Dr. Coffin explained, “people need opioids from somewhere. So when they can’t get them
`
`from a prescription, they end up going to the street.” Coffin, May 26, 2022, Trial Tr. at
`
`1900:3–7; see also Lembke Decl. ¶ 156. In addition, people suffering from more severe
`
`opioid addiction will seek out “more potent, plentiful, and cheaper forms over time.”
`
`Lembke Decl. ¶ 156. The foreseeable result is increased rates of heroin use and heroin
`
`overdose deaths.4 See Keyes Decl. ¶¶ 108, 123–26; Coffin ¶¶ 24–25.
`
`
`4 Defendants’ expert Douglas Tucker testified that he was not aware of any evidence that
`prescription opioid use serves as a “gateway” to heroin or fentanyl use. See Tucker (dkt. 1423) ¶
`40. In light of the significant and credible evidence offered at trial demonstrating that prescription
`opioid use is correlated with heroin and fentanyl use, the Court assigns Dr. Tucker’s testimony on
`11
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 12 of 112
`
`
`
`In the context of opioid addiction, the gateway effect similarly describes the
`
`tendency of people suffering from opioid addiction to use stronger forms of opioids over
`
`time. See Lembke Decl. ¶ 156. After developing a tolerance to prescription opioids, the
`
`“natural progression” for opioid users is to transition to more powerful opioids, like heroin.
`
`See id. Dr. Colwell, Chief of Emergency Medicine at Zuckerberg San Francisco General
`
`Hospital (“ZSFG”), testified that “approximately two-thirds of the patients who present to
`
`the [emergency department] with an opioid-related medical condition report that their
`
`addiction started with pills.” Colwell Decl. ¶ 10. Many of Dr. Colwell’s opioid-addicted
`
`patients “can trace very specifically their opioid use disorder and their addiction to a
`
`specific event.” Colwell, April 28, 2022, Trial Tr. at 358:7–22. Dr. Colwell explained:
`
`“Usually it’s an injury, a traumatic injury, or a surgery where they were prescribed
`opioid pills, and this starts a process or a pattern that either results in their feeling
`like those pills are no longer helping their pain and they need to seek more
`powerful, more potent ways of managing those or they have [gotten] to the point
`where physicians have tried to wean them off and they don’t—they’re no longer
`feeling like they can do that. And so they then seek other areas, other ways to—to
`manage their pain, to manage their—what they’re suffering from.”
`
`Id.
`
`
`
`The association between prescription opioid use and heroin use is even stronger for
`
`people who use prescription opioids non-medically. Taking prescription opioids non-
`
`medically makes the likelihood of using heroin significantly higher. See Keyes Decl. ¶¶
`
`60–63. One study found that individuals using prescription opioids non-medically have a
`
`5 times higher likelihood of using heroin than individuals who do not use prescription
`
`opioids non-medically. Id. ¶ 61. A 2020 longitudinal study found that individuals who
`
`used opioids medically were 2.7 times more likely to transition to heroin compared to
`
`individuals who did not use opioids. Id. ¶ 63. Individuals who used opioids non-
`
`medically were 6.8 times more likely to transition to heroin. Id. A different 2020 study
`
`found that only 9.5% of individuals who began using heroin had “never used
`
`this subject no weight.
`
`
`
`12
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`Northern District of California
`
`United States District Court
`
`

`

`Case 3:18-cv-07591-CRB Document 1578 Filed 08/10/22 Page 13 of 112
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`pharmaceutical opioids in a non-prescribed manner” before beginning heroin use. Id. ¶ 64.
`
`Consistent with these studies, an internal Walgreens presentation reported that “[n]early
`
`80% of Americans using heroin reported misusing opioids first” and that “[i]ndividuals
`
`who misuse prescription opioid pain pills are forty times more likely to abuse heroin.” P-
`
`25545 at 00008. The increased rates of heroin use in San Francisco in the 2010s reflect the
`
`link between prescription opioids and heroin use.
`
`c. Wave Three: Prescription Opioids, Heroin, and Fentanyl
`
`Fentanyl is cheap to produce, and it is extremely potent. Coffin Decl. ¶ 35. In the
`
`mid- 2010s, drug traffickers responded to increasing demand for opioids by increasing the
`
`production and distribution of black-market fentanyl. Id. ¶ 36. Fentanyl emerged on the
`
`East Coast around 2013, and it arrived in San Francisco around 2015. Coffin, May 26,
`
`2022, Trial Tr. at 1900:8–11. The lowe